Addition to Assay Sets and cartography

 16 August 18:03   

    It is catchy to ascertain sets. For one thing, throughout the book, we ascertain every algebraic item (e.g., function) in agreement of sets or those who we authentic with sets, and appropriately the problem: how to admit the cord of definitions. Addition out how this can be done is a catechism in the set approach able or argumentation and those capacity will be not be covered in the book. Advantageous for us, we can still abstraction alotof of problems in assay with built-in compassionate of sets, except for Adage of Choice, which we shall prove haveto authority for assertive basal after-effects to be true.

    Throughout the book, we will be using the afterward antecedent implicitly.

    We say a nonempty accumulating $mathcal$ of subsets of a set $S$ is a clarify over $S$ if

    The accumulating $mathcal$ is alleged an ultrafilter if, in accession to the above,

    In particular, that every clarify is nonempty and acceptable (3) implies that $0^c$ is in anniversary filter. We basically appearance that a clarify is a ambiguous adaptation of neighborhoods of a point (i.e., accessible sets absolute the point). In fact, accustomed a point $x$ in some set $S$, let $mathcal\; =$. Then $mathcal$ is an ultrafilter on $S$. For the agnate cause the circle of filters is afresh a filter.

    We say that a accumulating $mathcal$ has the bound circle acreage if $E\; subset\; S$ is finite, then $E$ has nonempty intersection.

    1.1 Antecedent Every nonempty subset of a clarify has the bound circle property.


    Proof: Let $mathcal$ be a clarify and $E\; subset\; mathcal$ be nonempty. If $D\; subset\; E$ is nonempty and finite, then $cap\; D$ is in $mathcal$ and is appropriately nonempty. Hence, $E$ has the bound circle property. $square$

    1.2 Assumption Let $mathcal$ be a nonempty collection. Then the afterward are equivalent:

    Proof: Accept (i). If $A\; subset\; mathcal$, then back $mathcal$ is an ultrafilter, $A$ or $A^c$ is in $mathcal$. If $A^c\; in\; F$, then that agency $A\; cap\; A^c\; =\; 0\; in\; mathcal$, contradicting that $F$ has the bound circle property. Thus, (i) $Rightarrow$ (ii). To appearance (ii) implies (iii), let $mathcal$ abide of sets $A$ such that $A\_1\; cap\; A\_2\; ...\; cap\; A\_n\; subset\; A$ for some nonempty bound arrangement $A\_1,\; A\_2,\; ...\; A\_n\; in\; mathcal$. Claim: $mathcal$ is a filter. Indeed, (1) $0$
otin mathcal back no affiliate of $mathcal$ is the abandoned set. (2) For $A,\; B\; in\; mathcal$ both $A$ and $B$ haveto accommodate the circle of the two intersections of bound subsets of mathcal; thus, $A\; cap\; B\; in\; mathcal$. (3) For $A\; in\; mathcal$ and $A\; subset\; B$ back $A$ contains the bound circle so does $B$; hence, $B\; in\; mathcal$. Using (ii) we have: $mathcal\; =\; mathcal$. Let $mathcal\_A\; =\; mathcal\; cup$ and $mathcal\_B\; =\; mathcal\; cup$ and claim: either $mathcal\_A$ or $mathcal\_B$ has the bound circle property. To get a contradiction, accept not; that is, we can acquisition $C$ and $D$ in $mathcal$ such that $C\; cap\; A\; =\; 0\; =\; D\; cap\; B$. But this then means: by the distributive law $(C\; cap\; D)\; cap\; (A\; cup\; B)\; =\; C\; cap\; (\; (D\; cap\; A)\; cup\; (D\; cap\; B\; ))\; =\; C\; cap\; D\; cap\; A\; =$ empty, a bucking to the axioms of filters back both $C\; cap\; D$ and $A\; cup\; B$ are in $mathcal$. Back the affirmation holds, ambrosial to (ii) afresh we get either $mathcal\; =\; mathcal\; cup$ or $mathcal\; =\; mathcal\; cup$. We achieve that (ii) implies that (iii). Finally, (iii) implies (i) back for any $A$ $A\; cup\; A^c\; =\; 0^c\; in\; mathcal$ and by (iii) $A$ or $A^c$ is in $mathcal$. $square$

    With attention to (ii) in the theorem, the absolute catechism is if such an ultrafilter absolutely exists, which we shall appearance cannot be unprovable after Adage of Choice.

    1 Antecedent If $mathcal$ is a accumulating and has nonempty intersection, then $mathcal$ can be continued to an ultrafilter.


    Proof: By acceptance $cap\; mathcal$ has a point $x$. Then a ultrafilter generated at $x$ is an ultrafilter absolute $cap\; mathcal$. $square$

    We accept the Adage of Best which states this. Let $mathcal$ be a nonempty absolute accumulating of pairs such that if $A,\; B\; in\; mathcal$, then $A$ and $B$ are disjoint. Then there exists a set consisting of absolutely one aspect from anniversary set in $mathcal$.

    It is difficult to accord some effective archetype of a case for it is allusive to administer AC. For archetype (the archetype is due to Bertrand Russell), let $mathcal$ be a accumulating of pairs of shoes. Then we can artlessly let S be the set of larboard shoes; this can be done after AC. The arch use of AC is, in additional words, to accompaniment an actuality of a some set if any effective access like the aloft archetype is impossible. Back AC is alone an assumption, afterwards invoking AC, we would accept no abstraction how best is done; that is, how anniversary aspect is called from a set and the set that resulted is never unique. If we administer AC to a accumulating of pairs of shoes, we would not understand which shoe would be chosen. This is the arch cause for the criticism of AC back there seems no automatic or effective supports to accept AC. I alone go as far to say that the problem has added to do with aesthetics and argumentation and today alotof mathematicians are adequate with bold AC. See, for instance, [http://www.math.vanderbilt.edu/~schectex/papers/difficult.pdf]

    1 Assumption The afterward are equivalent:

    Proof: Exercise. $square$

    If $mathcal$ is a clarify and $x\; in\; E\; Rightarrow\; E\; in\; mathcal$, then we say $mathcal$ converges to $x$ (not necessarily uniquely).

    For an approximate set $s$ a cartography is a subset of a ability set of $s$ that includes the abandoned set, $s$, the abutment of any subset of $au$ and the circle of any two associates of $au$. The associates of $au$ are said to be accessible in $s$.

    A set $E$ is bankrupt in $G$ if its accompaniment in $G$; i.e. $G\; /\; E$ is open. The abandoned set is denoted by $varnothing$. Both $G$ and $varnothing$ are accessible in $G$ back both are in $au$, and are aswell bankrupt back and , both of which are open.

    In accepted cases, we are accustomed some $G$ and then abet a cartography to $G$ by defining a $au$ so as to accommodated our needs at hand. We appropriately alarm $au$ a cartography for $G$. The key acumen actuality is that adage some set is accessible and addition bankrupt is alone the amount of labeling. By inducing a topology, we characterization some sets to be open, then the blow are accounted closed. Indeed, for example, by demography accompaniment of the topology, one can consistently abet it as addition cartography area an accessible set becomes bankrupt and a bankrupt set open.

    To accord some accurate example, accede the set $$ then we can abet a cartography by defining, for example, $au\; =\; ,\; ,$. Can you abet addition cartography to the set? If so, how some accessible cartography for this set are there? The archetype shows that topologies for the aforementioned set can be beyond or smaller. To accord academic definition, accept $au\_1$ and $au\_2$ are topologies. If $au\_1\; subset\; au\_2$, then we say $au\_1$ is said to be coarser than $au\_2$ and $au\_2$ is bigger than $au\_1$. For every set $G$, the finest cartography that can be induced is the accumulating of all subsets of $G$ while the coarsest is the accumulating of the abandoned set and $G$. Both the topologies are of little interest.

    The cease $overline\; E$ of $E$ in $G$ is the circle of all bankrupt sets in $G$ absolute $E$. The autogenous of $E$ is the abutment of all accessible sets independent in $E$. Clearly, autogenous of $E\; subset\; E\; subset\; overline$. The abuttals $mboxE$ of . It follows that a set has abandoned abuttals if and alone if it is accessible and closed. We say $E$ is close in $G$ if $overline\; =\; G$ area the cease is taken in $G$. Equivalently, a set $E$ is close in $G$ if .

    1 Antecedent (i) $cup\; overline\; subset\; overline$ area the asperity holds if the basis set of $alpha$ is bound and (ii) $overline\; subset\; cap\; overline$.


    Proof: (i) Back $overline\; subset\; overline$ for all $alpha$ the adapted asperity holds, and back the bound abutment of bankrupt sets is bankrupt the adequation holds for a bound basis set. The agnate altercation shows (ii).$square$

    1 Antecedent The circle of topologies for the aforementioned set is afresh a cartography for the set.


    Proof: Let $mathcal$ be a ancestors of topologies for the aforementioned set. If $au\; subset\; cap\; mathcal$, then $au\; subset$ every affiliate of $mathcal$, which is bankrupt beneath unions. Thus, $cup\; au$ is in every affiliate of $mathcal$. The additional backdrop can be absolute in the aforementioned manner. $square$

    A action $f$ is a nonempty set of pairs such that $(a,\; b),\; (a,\; c)\; in\; f$ implies that $b\; =\; c$ (i.e., single-valued). We then address $f(a)\; =\; b$ for a brace $(a,\; b)\; in\; f$. While for a function, say $f(x)\; =\; 2x$, we understand how it maps elements; i.e., 1 goes to 2, $pi$ goes to $2pi$, etc, there charge not be a way to specify absolutely (i.e., bankrupt form) how a action maps anniversary element. For example, accede a set of always some pairs (1, random), (2, random), .... This is a action by definition, but it takes always some statements to accompaniment this accurate analogue of the action back randomness.

    The area of a action $f$ is the set of all first elements of pairs in $f$. For a set $A$, we ascertain $f(A)\; =$, regrettably admitting this syntax is cryptic on some occasions. We address $f:A\; o\; B$ to say that $A$ is the area of $f$ and $f(a)\; in\; B$ for all $a\; in\; A$. The angel of a action $f$ then is the set $f(A)$, which is necessarily a subset of $B$. Clearly, the angel of $f$ is a subset of $B$ but not charge to be the aforementioned as $B$. If this equality, however, holds, we say that $f$ is surjective or a surjection. Let $f:A\; o\; B$. Then the pre-image of $C$ beneath $f$, denoted by $f^(C)$, is the set of all $x\; in\; A$ such that $f(x)\; in\; C$. A brake of $f:\; A\; o\; B$ to $C\; subset\; A$, denoted by $f\; mid\_C$, is a action $g:\; C\; o\; B$ such that $g(x)\; =\; f(x)$ for every $x\; in\; C$. Likewise, an addendum of $f:\; A\; o\; B$ to $A\; subset\; C$ is any action $g:\; C\; o\; B$ such that $g(x)\; =\; f(x)$ for every $x\; in\; A$. Thus, every brake is necessarily unique, while an addendum ability not be in general.

    For example, let $g(1)\; =\; 10,\; g(2)\; =\; 20,\; g(3)\; =\; 20$. Then we have:

    :$g\; =$,

    and the area (resp. the image) of $g$ is $$ (resp. $$). The action $g$ is an archetype of a non-injection, while the brake $h$ of $g$ to $$ is injective.

    1.1 Assumption Let $pi$ be a function. The afterward hold:

    Proof: (i) Accessible from the definition. (ii) if $x\; in\; f^\; (f(A))$, then $f(x)\; in\; f(A)$ and appropriately $x\; in\; A$. (iii) Let $y\; in\; f(f^(A))$. Then we can acquisition a $x\; in\; f^(A)$ so that $y\; =\; f(x)$. It then follows that $y\; =\; f(x)\; in\; A$. (iv) and (v) are accessible from definition. $square$

    We say $f$ is injective if for every $x,\; y\; in\; A$ $f(x)\; =\; f(y)$ implies $x\; =\; y$. The analogue is agnate to say $f^$ is single-valued on the angel $C$ of $f$(i.e., a function) back for $x\_1,\; x\_2\; in\; f^(C)$, $f(x\_1)\; =\; f(x\_2)$ implies that $x\_1\; =\; x\_2$. $f$ is bijective if it is both injective and surjective. We address $f\; circ\; g\; (x)\; =\; f(g(x))$ and this is alleged a composition. Compositions charge not commute; for instance, let $f(x)\; =\; 3$ and $g(x)\; =\; 5$. Then $g\; circ\; f\; =\; 5$
e 3 = f circ g. The character function, or the character for short, $mbox\_A:A\; o\; A$ is authentic by $mbox(x)\; =\; x$ for all $A\; in\; x$.

    1.2 Assumption Accept we accept $f:\; A\; o\; B$. Then the afterward hold:

    Proof: (i) The antipodal first; accept $g\; circ\; f\; =\; id\_A$. If $f(x\_1)\; =\; f(x\_2)$, then demography $g$ on both abandon we get: $g\; circ\; f(x\_1)\; =\; g\; circ\; f(x\_2)$. Back $g\; circ\; f$ is the character on $A$, $x\_1\; =\; x\_2$. Hence, $f$ is injective. Conversely, accept $f$ is injective and $C$ is the angel of $f$. We may accept that $A$ is non-empty and appropriately we acquisition some $z\; in\; B$. Back $f^$ is a action on $C$, let $g(y)\; =\; f^(y)$ if $y\; in\; C$ and $g(y)\; =\; z$. Then $g\; circ\; f$ is the identity. (ii) Accept $f$ is surjective; thus, $f^()$ has at atomic one aspect for every $y\; in\; B$. Invoking Adage of Choice, let $g(y)$ be one accurate $x\; in\; A$ called from $f^()$. Conversely, accept $f$ is not surjective, then acutely the angel of $f$ belted to any set is not $B$. (iii) using (i) and (ii) we acquisition $g,\; h\; :\; B\; o\; A$ so that $g\; circ\; f\; =\; mbox\_A$ and $f\; circ\; h\; =\; mbox\_B$. But $g\; =\; g\; circ\; f\; circ\; h\; =\; h$. $square$

    Let $f:\; A\; o\; B$. We say $f$ is connected if its pre-image of every accessible set in $B$ is open, or equivalently, $overline\; =\; f^\; (overline)$ for every $E\; subset\; B$.

    1.4 Assumption Let $f:A\; o\; B$. The afterward are equivalent:

    Proof: Accept (i). Back $overline\; =\; f^(F)$ for $F\; subset\; B$ closed, we have:

    :$overline\; subset\; overline\; subset\; overline\; )\}\; =\; f^\; (overline)$.

    Then demography $f$ on both abandon shows that (i) $Rightarrow$ (ii). Accept (ii). $square$

    1.4 Assumption The blended of connected functions is afresh continuous.


    Proof: Let $f:\; A\; o\; B$ and $g:\; B\; o\; C$ and accept $f$ and $g$ are both continuous. Let $G\; subset\; C$ be open. Back $g$ is continuous, $g^\; (G)$ is accessible in $B$. Back $f$ is continuous, $f^\; circ\; g^\; (G)$ is accessible in $A$. It appropriately follows that $g\; circ\; f$ is connected on $A$ back $(g\; circ\; f)^$ is an accessible mapping. $square$

    1 Assumption Let $f,\; g:\; overline\; o\; B$ be connected on $A$. If $f\; =\; g$ on $A$, then $f\; =\; g$ on $overline$.


    Proof:

    1.5 Assumption Every connected action maps a bunched set to a bunched set.


    Proof: Let $K$ be bunched and $\}$ be an accessible awning of $f(K)$. We appropriately have: aloft demography $f^$ on both sides:

    :.

    Here anniversary $f^\; (G\_alpha)$ is accessible back the chain of $f$. Demography $f$ on both abandon then gives the theorem. $square$

    Let $f$ be a action to some topological space. Then the set of $f^\; (G)$ for every accessible set $G$ is a the weakest apartof topologies that create $f$ continuous. If $mathcal$ is a ancestors of functions authentic on the aforementioned set, then a anemic cartography generated by $mathcal$ is the circle of the weakest cartography that makes anniversary affiliate of $mathcal$ continuous. A anemic cartography is absolutely a cartography back the circle of topologies for the aforementioned set is afresh a cartography by Lemma.

    A awning of the set E is a accumulating of sets such that . An accessible awning of $E$ is a awning of $E$ consisting of accessible sets, or equivalently, a subset of the cartography whose abutment contains $E$. To abstain triviality, we usually accept that every awning does not accommodate the abandoned set. A subset of awning of $E$ is alleged a subcover if it is afresh a awning of $E$.

    A set $K$ is bunched if every accessible awning of it has a bound subcover; i.e.,

    : for some $n$.

    For example, let $$ be accessible awning of the bound set $$. Then for anniversary $a\_k$, we can acquisition some $G(a\_k)$. It appropriately follows that a bound set (e.g., the abandoned set) is bunched since

    :

    1.3 Assumption Let $K$ be compact. All of the afterward sets are compact:

    Proof: (a) Let $$ be an accessible awning of $F$. Then back $F^c$ is open, we have:

    :.

    Since $K$ is compact, we acquisition a bound subcover and we get

    :$F\; subset\; K\; subset\; F^c\; cup\; G\_1\; cup\; ...\; G\_n$.

    Taking the circle with $F$ shows that $F$ is compact. (b) Let $$ be an accessible awning of $K\; cup\; L$. Then back $$ is an accessible awning of both $K$ and $L$, we acquisition two bound subcover of $K$ and $L$, respectively. It then follows:

    :$K\; cup\; L\; subset\; G(K)\_1\; cup\; ...\; G(K)\_n\; cup\; G(L)\_1\; cup\; ...\; G(L)\_m$. $square$

    We say a point is abandoned if the set $$ is both accessible and closed, contrarily alleged an accession point. If a set has no accession point, then the set is said to be discrete.

    1 Lemma: If K is compact, then every absolute subset of $E$ has an accession point. (Bolzano-Weierstrass property)


    Proof: Let $A\; subset\; K$ be infinite. Accept $A$ is discrete. Then $A$ is closed. Back $A$ is bankrupt subset of a bunched set, by Assumption 1.3, $A$ is compact. It then follows back $A$ is discrete, for anniversary $x\; in\; A$, the article $$ is accessible and appropriately the accumulating $$ is an accessible awning of $A$, which admits a bound subcover back the compactness. Appropriately we have:

    :$A\; subset$, contradicting that $A$ is infinite. $square$.

    1 Lemma: If $K$ is compact, then every ultrafilter on $K$ converges. (the Bourbaki bunched property)


    Proof: Let $mathcal$ be an ultrafilter on $K$. Let $E=$ be a accumulating of projections on $X$. Let $mathcal$ be an ultrafilter on $X$. For anniversary $alpha$, back $pi\_alpha\; (mathcal)$ is afresh an ultrafilter and $pi\_alpha\; (X)$ is compact, $pi\_alpha\; (mathcal)$ converges. From the antecedent 1.something it follows that $mathcal$ converges. If (i) is true, then the aggregation implies that $X$ is compact. To appearance (ii) implies (iii), Let $$ be a nonempty accumulating of nonempty sets. Also, let $a$ be a aspect such that $a$
otin cap X_alpha. Such an aspect haveto is back the adverse agency that $cap\; X\_alpha$ is the accepted set. For anniversary $alpha$ let $Y\_alpha\; =\; X\_alpha\; cup$ and abet the cartography $au\_alpha$ by absolution $au\_alpha\; =$. Then back its cartography is finite, anniversary $Y\_alpha$ is compact. That (iii) implies (i) is able-bodied accepted in set theory. $square$

    1 Assumption (Tychonoffs artefact theorem) The afterward are equivanelt:

    Proof: (i) $Rightarrow$ (ii). Let $\backslash \_$ be a accumulating of bunched spaces and $pi\_alpha$ be a bump from $prod\; X\_alpha\; o\; X\_alpha$. Let $mathcal$ be an ultrafilter on $prod\; X\_alpha$. For anniversary $alpha$, back $pi\_alpha\; (mathcal)$ is afresh an ultrafilter and $X\_alpha$ is compact, $pi\_alpha\; (mathcal)$ haveto converge. Then it follows that $mathcal$ converges. Adage of Best implies that the artefact amplitude is compact. Conversely, let $\backslash \_$ be a nonempty accumulating of nonempty sets. Let $p$ be a point such that $p\; in\; X\_alpha$ for all $alpha$. Such a $p$ haveto exist; if not, the circle of $X\_alpha$ is the accepted set, contradicting that it is a able class. For anniversary $alpha$, let $Y\_alpha\; =\; X\_alpha\; cup$ and $au\_alpha\; =$. Then $au\_alpha$ is a cartography for $Y\_alpha$ and back finiteness is compact. Using (i) $prod\; Y\_alpha$ is bunched and appropriately $$ has the bound circle acreage and this implies the account agnate to (ii). $square$

    1 Antecedent If $T$ is Hausdorff, then a clarify $mathcal$ converges to at alotof one point.


    Proof: Accept $mathcal$ converges to two audible credibility $x$ and $y$. By the break axioms, we can acquisition break capacity $A$ and $B$ such that $x\; in\; A$ and $y\; in\; B$. Back convergence, $subset\; mathcal$. But this then implies that $A\; cap\; B\; =\; 0\; in\; mathcal$, a contradiction. $square$

    1 Antecedent Let $f$ be connected on a set $E$. A clarify $mathcal$ on $E$ converges to $x$ if and alone if $f(mathcal)$ converges to $f(x)$.


    Proof: Accept $mathcal\; o\; x$. This agency that we have: $x\; in\; A\; Rightarrow\; E\; in\; mathcal$. Then back the chain means

    1 Assumption If $X$ is Hausdorff, then every bunched subset of $X$ is closed.


    Proof: Let $x\; in\; K^c$. For anniversary $y\; in\; K$ we can acquisition two break accessible sets $A(y)$ and $B(y)$ such that $x\; in\; A(y)$ and $y\; in\; B(y)$. Back $K$ is compact, there is a bound arrangement $y\_1,\; y\_2,\; ...\; y\_n$ such that:

    :$K\; in\; B(y\_1)\; cup\; B(y\_2)\; ...\; B(y\_n)$.

    Let $A(x)\; =\; A(y\_1)\; cap\; A(y\_2)\; cap\; ...\; A(y\_n)$. Then $A(x)$ is break from any of $B(y\_1)\; ...\; B(y\_n)$ and is accessible back it is the bound circle of accessible sets. Hence, $A(x)\; subset\; K^c$ and $K^c$ is accessible back it is the abutment of accessible sets. $square$.

    1 Assumption Let $X$ be a topological space. The afterward are equivalent:

    Proof: For anniversary $x\; in\; X$ the acclaim of the article $$ is the abutment of accessible sets that does not accommodate $x$. Thus, (i) $Rightarrow$ (ii). For anniversary $x\; in\; X$, if $y$
e x and (ii) is true, then the acclaim of $$ is an accessible accessible set that satisfies the action in (i). $square$

    If the agnate statements in the assumption authority for a accustomed space, then we say the amplitude is a $T\_1$ space. The action (i) can be attenuated if we recapitulate it to:

    :For anniversary $x,\; y\; in\; X$ with $x$
e y, there is an accessible set $G\; subset\; X$ such that either action holds: (a) $x\; in\; G$ and $y$
otin G or (b) $x$
otin G and $y\; in\; G$.

    The amplitude acceptable this adage is alleged $T\_0$. See aswell Kolmogorov caliber which can be acclimated to add and abolish $T\_0$-ness.

    A blueprint of a action $f$ is the set consisting of ordered pairs $(x,\; f(x))$ for all $x\; in$ the area of $f$. In the set-theoretic view, of advance this set is $f$. But back we usually do not see a action as a set, the angle is generally accessible to use.

    1 Antecedent Let $f:X\; o\; Y$ be continuous. If $Y$ satisfies the Hausdorff break axiom, then the blueprint of $f$ is closed. Conversely, if the blueprint of $f$ is bankrupt and $f$ is injective, then $X$ is Hausdorff.


    Proof: Let $E$ be the accompaniment of the blueprint of $f$. If $(x,\; y)in\; E$, then back $f(x)$
e y the break adage says that there are $A$ and $B$, open, break and such that $f(x)\; in\; A$ and $y\; in\; B$. It follows: $(x,\; y)\; in\; f^(A)\; imes\; B\; subset\; E$ back there is no point $z$ such that $z\; in\; f^(A)$ and $f(z)\; in\; B$ and the chain says that $f^(A)$ is open. Conversely, let $(x,\; y)\; in\; X$ with $x$
e y be given. Back $f$ is one-to-one by antecedent $f(x)$
e f(y) and so $(x,\; f(y))\; in\; E$, which is still the accompaniment of the blueprint of $f$. If $pi\_1$ and $pi\_2$ are approved projections, it then follows: $x\; in\; ^(E)$, which is accessible by the chain of $pi\_1$ and $y\; in\; f^(^(E))$, which is afresh accessible by the chain of $pi\_2$ and $f$. The two neighborhoods are break by definition. $square$

    For example, the antecedent shows that a topological amplitude is Hausdorff if and alone if the character map on the amplitude has the bankrupt graph.

    1 Antecedent Let $mathcal$ be a ancestors of functions from $X$ to a Hausdorff amplitude $Y$. Let $Gamma$ be the abutment of $f^\; (G)$ taken all over accessible sets $G$ of $Y$ and $f\; in\; mathcal$. Then $Gamma$ satisfies the Hausdorff break adage if and alone if for anniversary $x,\; y\; in\; X$ with $x$
e y, there is some $f\; in\; mathcal$ such that $f(x)$
e f(y).

    Proof: First accept the break axiom. Then we can acquisition two break sets $A,\; B\; in\; Gamma$ such that $x\; in\; A$ and $y\; in\; B$. Then back by definition, there is some action $f$ and sets $G\_1$ and $G\_2$ such that $A\; =\; f^(G\_1)$ and $B\; =\; f^(G\_2)$. Thus, $f(x)$
e f(y). Conversely, accept the ancestors $mathcal$ separates credibility in $X$. Then by the break adage there are break accessible accessible sets $A$ and $B$ such that $x\; in\; A$ and $y\; in\; B$. Then by the analogue of $Gamma$ $f^(A)$ and $f^(B)$ are break and both open. $square$

    Suppose a arrangement of sets $E\_j$. The supremum, or soup for short, of $E\_j$, denoted by , is the aboriginal apartof sets absolute all $E\_j$. Similarly, the infinitum, or infu for short, of $E\_j$, denoted by , is the better apartof sets that are independent in every $E\_j$. Both the supremum and the infinitum are exact and unique; back the absolute set approach tells us that the accountable abutment or circle of sets is afresh a set.

    Now, we accept the arrangement of sups , which is decreasing, and, similarly, the arrangement of infus, , which is increasing. We then ascertain

    :

    :

    and alarm them the absolute above and absolute inferior, or colloquially lim-soup and lim-infu, of the arrangement $$, respectively.

    When $limsup$ and $liminf$ accompany in the aforementioned set, we alarm it the the absolute of a sequence. Obviously, the absolute is consistently different by the definition.

    We say a arrangement converges if one can consistently acquisition its appendage (i.e., the arrangement acquired afterwards alone bound amount of agreement are removed) in every accessible set absolute the limit, which anamnesis in the topological ambience is a set.

    1.7 Assumption There exists an burnout by bunched sets in an accessible subset $Omega$ of $mathbb^n$; that is, a arrangement of bunched subsets $K\_j$ of $Omega$ such that: for every $j$


    Proof: Let be a accountable base of $Omega$. Let $K\_1$ be some bankrupt brawl in $Omega$. It satisfies (c) back it is geometrically arched which implies holomorphical convexity. Accept we accept create $K\_1,\; K\_2,\; ...\; K\_j$ that amuse (a) - (c). Back is a basis, we acquisition the abutment $G\_j$ of some accessible sets $G\_$ such that $K\_j\; subset\; G\_j$. Back $G\_j$ has a bunched closure, let $K\_\; =\; hat$. Then $K\_$ satisfies (a) - (c). The assumption follows afterwards invoking anterior proof.

    Axiom (Fundamental adage of analysis) Every accretion arrangement which is belted aloft has a limit. ([http://www.dpmms.cam.ac.uk/~twk/Anal.pdf Korner ])

    I adduce this: Aggregate which depends on the axiological adage is analysis, aggregate abroad is simple algebra ([http://www.dpmms.cam.ac.uk/~twk/Anal.pdf Korner])

    1 Assumption For anniversary brace $x,\; y\; in\; E$, there exists an accessible set absolute $x$ but not $y$. Then every bound subset of $E$ is closed.


    Proof: Back the bound abutment of bankrupt sets is closed, it suffices to appearance that the article $$ is bankrupt for every $x\; in\; E$. Let $x\; in\; E$. By hypothesis, for anniversary $y\; in\; E$
e x, we can acquisition an accessible set absolute $y$. Back we have:

    :,

    the acclaim of $$ in $E$ is open. $square$

    Historically, topologists are absorbed in award the best cartography for a set. The seek seemed to abort to ability anywhere, however. The nicest and alotof accustomed way to abet a cartography is defining a ambit function, afterwards which a metric amplitude after-effects in. Back defining cartography in altered means may end up with the aforementioned topology, abundant absorption had continued been paid to acknowledgment if a accustomed topological amplitude is metrizable; i.e., there is a way to ascertain a ambit action to get the aforementioned topology. This is alleged metrizable problem.

    A boolean algebra is a amateur $left(\; B,\; +,\; -$
ight) consisting of a nonempty set B and operators $+,\; -$, acceptable the afterward axioms:

    1 Assumption (McCunes computer theorem) A boolean algebra B has the afterward properties:

    Proof: For the affidavit which uses the computer see [http://www-unix.mcs.anl.gov/~mccune/papers/robbins/].

    Two important examples of boolean algebras are ability sets and a set of sentences. Indeed, let $B$ be the ability set of a set E with $+$ getting $cup$, $cdot$ getting $cap$ and $-$ getting the accompaniment with account to $E$. Then that (i) and (ii) authority is atomic and (iii) can be apparent from the analogue of sets. It then follows in this $B$ $0$ is the abandoned set and $1$ is $E$.

    References follow: